The aim of this study was to investigate if dynamic gadolinium-DTPA-supported magnetic resonance (MR) imaging can monitor the therapeutic effect of a fast-acting immuno-modulating drug like anti-tumour necrosis factor alpha (anti-TNF-alpha) monoclonal antibody (moab) in patients with rheumatoid arthritis (RA). Dynamic MR imaging was performed on 64 joints in a total of 18 patients before and after infusion with either a placebo or 1 or 10 mg/kg of anti-TNF-alpha moab. Additionally, treating the placebo group and reinfusing the verum group with either 3 or 10 mg/kg was monitored by quantitative nuclear magnetic resonance (NMR). Time-dependent signal intensity changes were then correlated with a total of five Paulus criteria and with ESR and C-reactive protein (CRP). No changes in either the gadolinium uptake or clinical parameters were seen after the infusion of a placebo. Therapy with 1 mg/kg anti-TNF-alpha moab resulted in a significant decrease in clinical disease activity, as well as in gadolinium-DTPA uptake in dynamic NMR studies. However, correlations between singal intensity changes and Paulus criteria were only demonstrated for the variable "doctor's evaluation of disease activity". Patients given 10 mg/kg moab demonstrated a very significant improvement in all clinical manifestations of their disease, as well as a high significant reduction in gadolinium uptake (P = 0.004). In addition, the latter group showed significant correlations between time-dependent signal intensity changes and five Paulus criteria: "number of swollen joints", "number of painful joints", "duration of morning stiffness", "doctor's evaluation of disease activity" and "patient's evaluation of disease activity". No differences and correlations were seen for ESR and CRP. We concluded that dynamic NMR studies are suitable to monitor inflammatory activity in RA patients under therapy with biological response modifiers such as anti-TNF-alpha moab.